Skip to content

Latest commit

 

History

History
167 lines (112 loc) · 5.13 KB

README.md

File metadata and controls

167 lines (112 loc) · 5.13 KB

alt text alt text

Pykson: A JSON Serializer/Deserializer for Python

Pykson is a JSON serializer/deserializer in python.

Tested with:

  • Python 3.6+

Use the following command to install using pip:

pip install pykson

Usage example

Create Object Models

First, create your object model which extends JsonObject

from pykson import JsonObject, IntegerField, StringField, ObjectListField


class Course(JsonObject):
    name = StringField()
    teacher = StringField()


class Score(JsonObject):
    score = IntegerField()
    course = Course()


class Student(JsonObject):

    first_name = StringField()
    last_name = StringField()
    age = IntegerField()
    scores = ObjectListField(Score)

Deserialize json strings

Use Pykson class to deserialize json string to JsonObjects

from pykson import Pykson

json_text = '{"first_name":"John", "last_name":"Smith", "age": 25, "scores": [ {"course": {"name": "Algebra", "teacher" :"Mr. Schmidt"}, "score": 100}, {"course": {"name": "Statistics", "teacher": "Mrs. Lee"}, "score": 90} ]}'
student = Pykson().from_json(json_text, Student)

Serialize objects

Use Pykson class to serialize JsonObjects to string

Pykson().to_json(student)

Fields

There are different types of predefined fields: IntegerField, FloatField, BooleanField, StringField, ListField, ObjectField, ObjectListField, DateField, TimeField, DateTimeField, TimestampSecondsField and TimestampMillisecondsField.

There are four other types of fields which help with storing fields with specific integer or string values. To create a field with multiple choice integer values, use MultipleChoiceIntegerField or EnumIntegerField classes. To create a field with multiple choice string values, use MultipleChoiceStringField or EnumStringField classes.

Example for MultipleChoiceStringField:

from pykson import MultipleChoiceStringField

class WeatherInfo(JsonObject):

  condition = MultipleChoiceStringField(options=['sunny','cloudy','rainy'], null=False)

Example for EnumStringField:

from enum import Enum
from pykson import EnumStringField

class WeatherCondition(Enum):
  SUNNY = 'sunny'
  CLOUDY = 'cloudy'
  RAINY = 'rainy'


class WeatherInfo(JsonObject):
  condition = EnumStringField(enum=WeatherCondition, null=False)

Advanced usage

Serialized names

It is possible to use change name of fields during serialization/deserialization. For this purpose, use serialized_name input in the fields

from pykson import Pykson, JsonObject, IntegerField, StringField, ObjectField
class Score(JsonObject):
    score = IntegerField(serialized_name="s")
    course = StringField(serialized_name="c")


class Student(JsonObject):

    first_name = StringField(serialized_name="fn")
    last_name = StringField(serialized_name="ln")
    age = IntegerField(serialized_name="a")
    score = ObjectField(Score, serialized_name="s")


json_text = '{"fn":"John", "ln":"Smith", "a": 25, "s": {"s": 100, "c":"Algebra"}}'
student = Pykson().from_json(json_text, Student)

Work with dates and datetimes

Pykson currenty has five fields for handling dates and datetimes. Three of them, DateField, TimeField and DateTimeField, use date/time formats to serialize/deserialize values. The other ones, TimestampSecondsField and TimestampMillisecondsField use integer values to serialize/deserialize datetimes.

Accept unknown key/value pairs when deserializing

from_json method currently has an input parameter named accept_unknown with default value of false. If you want to deserialize an string to a JsonObject and ignore unknown keys which are not defined in your model class as fields, you can set this parameter to true. If this parameter is false, an error is raised when facing an unknown key in the json.

json_text = '{"fn":"John", "ln":"Smith", "a": 25, "up":"some unknown parameter", "s": {"s": 100, "c":"Algebra"}}'
student = Pykson().from_json(json_text, Student, accept_unknown=True)

Type hierarchy adapter

You can register multiple type hierarchy adapters using register_type_hierarchy_adapter method of 'Pykson' class.

from pykson import TypeHierarchyAdapter

class Student(JsonObject):
    name = StringField(serialized_name="n")


class HighSchoolStudent(Student):
    high_school_name = StringField(serialized_name="sn")


class UniversityStudent(Student):
    university_name = StringField(serialized_name="un")


students = [
    HighSchoolStudent(name="john", high_school_name="Redstone High"),
    UniversityStudent(name="alice", university_name="Green Institute of Tech.")
]

pson = Pykson()
pson.register_type_hierarchy_adapter(
    Student,
    "student_type",
    {
        "highschool": HighSchoolStudent,
        "university": UniversityStudent
    }
)

students_json = pson.to_json(students)

decoded_students = pson.from_json(students_json, Student)

assert decoded_students == students